Ergonomics advice for developers is mostly a shopping list. The literature it implicitly claims to rest on does not support a shopping list.
An overview of systematic reviews on neck and upper-extremity disorders among computer users finds intervention evidence that is limited and inconsistent across studies. That is not a fringe reading — it is what happens when you try to run controlled trials on furniture. You cannot blind anyone to which chair they are sitting in, comfort is self-reported, and the people who volunteer for an ergonomics study tend to be the ones already convinced.
So this post is organised the other way round: what is free and probably matters, then what is cheap and might, then what to buy last.
Geometry, which costs nothing
Most bad setups are not bad because of what was bought. They are bad because things are at the wrong heights relative to each other, and nobody measured.
Element | Defensible position | What goes wrong |
|---|---|---|
Screen height | Top of the usable area near eye level | Laptop on a desk: months of neck flexion, all day |
Screen distance | Roughly arm's length, further for larger displays | Too close, on the assumption that closer is easier to read |
Elbow angle | Around 90 degrees or slightly open, shoulders down | Desk too high, so shoulders sit raised for hours |
Wrists | Roughly neutral, not extended upward | Keyboard tilted up on its feet, which is the default |
Feet | Flat, supported, thighs roughly horizontal | Dangling, because the chair was raised to fix the desk |
Mouse position | Close in, same plane as the keyboard | Out to the side, arm abducted and statically loaded |
Two of those rows do most of the damage in practice. The first is a laptop used on a desk without a stand — the screen is necessarily below eye level, so the neck is flexed for the whole working day, every day, and it is the single most common setup in this industry. For what an ergonomic workspace actually needs framed around revenue rather than architecture, XenGrowth is the better starting point.
The second is the mouse. The reviews find higher probability of upper-limb disorders in mouse-intensive work than keyboard-intensive work, and a mouse placed out to the side of a full-width keyboard holds the arm abducted and statically loaded for hours. Moving it closer costs nothing; a keyboard without a number pad, if you don't use one, is the cheapest genuinely evidence-adjacent purchase available here.
The laptop-on-a-desk setup and the mouse held out to one side are free to fix and are what most expensive chairs are bought to compensate for.
Variation, which is a behaviour rather than a purchase
The idea of a single correct posture held all day is not well supported, and it is a slightly odd goal on the face of it. Static loading is generally harder on tissue than varied loading, which is why a perfect position maintained for nine hours is still nine hours of one thing.
So the property worth designing for is how easily you can change. A monitor arm that moves, a chair whose adjustments you can reach without tools, a desk you can work at standing occasionally — these are worth having not because any one position is right but because switching is cheap. On the operations side of this specifically, The XenGrowth resource library is worth reading.
Meyer et al.'s instrumented data is the reason this needs to be easy rather than deliberate. Developers switched activity every 0.3 to 2.0 minutes outside meetings, and one participant coded for 135.7 minutes without a break longer than two. Nothing that requires remembering will survive that second pattern, which is the one worth interrupting.
Lighting and the second monitor
Two elements that rarely appear in ergonomics checklists and are worth a paragraph each, because both are geometry problems disguised as preference.
Lighting mostly matters through contrast rather than brightness. A bright screen in a dark room means the eye is repeatedly adapting between two very different luminance levels every time you look away, and a window directly behind the monitor produces the same problem continuously. The cheap fix is ambient light behind and around the screen at a level roughly comparable to the display, and never having a bright window in the field of view behind the monitor. None of this requires equipment; it usually requires turning the desk ninety degrees.
Multiple monitors introduce a geometry question nobody asks. Two screens of equal size placed side by side means the centre of your work is the bezel between them, so the neck spends the day rotated to one side or the other. If one screen carries most of your attention, that screen belongs directly in front of you and the second one goes off to the side for things you glance at — logs, chat, documentation. A symmetrical arrangement looks tidier and is worse, and this is one of the few places where the mechanism is uncontroversial even though the outcome evidence is thin. XenGrowth on AI agents and marketing automation works through AI agents and marketing automation in more operational detail.
Both of these follow the same rule as everything else here: they are free, they are about relationships between things rather than about the things themselves, and almost nobody has checked them. That combination is what makes geometry the right place to start and the shopping list the wrong one.
The spending order
If money is going to be spent, this is roughly the order the evidence and the contact hours suggest — noting that the evidence is weak throughout and the ordering is partly reasoning from mechanism.
A laptop stand and a separate keyboard, if you use a laptop as your main machine. This is the largest single geometric improvement available and it costs very little
A chair that fits your body, adjusted properly. Weak evidence, but you are in it for six or seven hours a day, and contact hours count for something even when the trials are inconclusive
A monitor arm, mostly because it makes position changes trivial rather than because any single height is correct
A keyboard without a number pad you don't use, so the mouse comes closer to the midline. Cheap, and aimed at the input the reviews are more suspicious of
A sit-stand desk, if and only if you will genuinely alternate. Otherwise it is an expensive fixed desk
Everything else — wrist rests, footrests, specialist mice, ergonomic keyboards. Not useless, but the evidence does not distinguish them, so buy them for comfort and stop justifying them as prevention
Wrist rests deserve a specific note because they are almost universal and the reasoning behind them is shaky. Resting the heel of the hand on a firm surface while typing creates sustained contact pressure over the carpal tunnel, which is not obviously what you want. They are intended as a place to rest between bursts rather than a support to type from, and almost nobody uses them that way.
What no workspace fixes
It's worth ending on the limit, because the whole category has an implicit promise that the right setup makes the problem go away.
The strongest evidence in this general area is not about workstations at all. Ekelund's harmonised meta-analysis of 1,005,791 people found that roughly 60 to 75 minutes a day of moderate-intensity activity appeared to eliminate the excess mortality risk associated with high sitting time. The WHO adds muscle-strengthening across all major muscle groups on two or more days a week. No chair produces either of those. XenGrowth on AI search, GEO and discovery goes further into AI search, GEO and discovery.
That is not an argument against a good setup. Comfort during six hours a day is worth having on its own terms, and geometry that stops your neck being flexed all afternoon is a genuine improvement. It is an argument against the substitution — the very common pattern where an expensive workspace is purchased and mentally filed as having dealt with the health question.
If you have | Fix this first | Cost |
|---|---|---|
A laptop on a desk | Stand plus external keyboard | Small, and it is the biggest single win |
Mouse out beyond a number pad | Move it in; consider a shorter keyboard | Zero to small |
Shoulders raised while typing | Lower the desk or raise the chair and support the feet | Zero |
Feet dangling | Anything solid under them | Zero |
Screen below eye level | Raise it, even on books | Zero |
None of the above | Stop reading ergonomics posts and go for a walk | Zero |
This is general information about published research, not medical or professional ergonomic advice, and persistent pain is a question for a clinician who can examine you rather than for a table of recommended angles.
The summary I would defend is this: the free changes are the ones most likely to matter, they are the ones almost everyone has not made, and the evidence base does not currently support anyone being confident about much beyond that.
Further reading from XenGrowth
The XenGrowth resource library — what you'll learn: how the commercial side of this work is run, across search, automation and revenue operations.
XenGrowth on AI agents and marketing automation — what you'll learn: how the teams who own AI agents and marketing automation plan and measure it.
XenGrowth on AI search, GEO and discovery — what you'll learn: how the teams who own AI search, GEO and discovery plan and measure it.
Where this work meets go-to-market
XenGrowth's growth engineering practice covers the go-to-market side of what an ergonomic workspace actually needs, which this piece deliberately leaves alone.
If you do exactly one thing after reading this, sit down at your desk and check the six rows of the geometry table against your actual setup. It takes five minutes, it costs nothing, and in most cases at least two of them are wrong in a way you have stopped noticing because you have been looking at it every day for years. That is the whole argument for measuring rather than buying: the defects you live with are precisely the ones you have adapted to seeing.
